AI-900 Azure AI Fundamentals | Azure Certification

Explore the features and use cases of common AI workloads on Azure, and learn how to select the appropriate Azure AI workloads for a given scenario on the AI-900 exam.

Explore AI 900 workloads from machine learning to Azure computer vision, including automated ML, Designer, notebooks, image classification, object detection, OCR, facial recognition, and video indexer.

Explore Azure AI and Azure Machine Learning services, including OpenAI for generative tasks, computer vision, speech recognition, image analysis, and natural language processing, with free and standard pricing tiers.

Explore the six guiding principles of responsible ai from Microsoft's standard - fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability - and how they shape ai design and deployment.

Explore Microsoft's six guiding principles for responsible AI—fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability—and how the Responsible AI Standard guides goals and actions for exam-ready understanding.

Identify how exam scenarios map use cases to responsible ai principles. Define the guiding principles as fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability.

Explore how data scientists prepare data, features and labels, and how training, validation, and test datasets build and evaluate machine learning models, including predictions and inferencing.

Explore regression, classification, and clustering as core machine learning techniques. Differentiate supervised from unsupervised learning and understand how binary and multiclass classification works.

Identify feature and label dependencies across supervised and unsupervised learning, distinguish regression, classification, and clustering with practical examples, and review the confusion matrix including true positives and true negatives.

Explore Azure machine learning capabilities including AutoML and Designer for building models, computing and data services, and deployment via endpoints for real-time, batch, and serverless use, including Azure OpenAI endpoint.

Provision an Azure Machine Learning Studio workspace, create a weather dataset from a CSV online, and explore notebooks, automated ML, and designers.

Learn common machine learning techniques in the Azure machine learning workspace, including regression, classification, time series forecasting, NLP, and computer vision, and explore automated ML jobs and designer modeling.

Train a regression model with automated ml on weather data. Predict temperature, use a weather job with 3 max trials on serverless compute, and evaluate normalized root mean squared error.

Explore exam tips for Azure Machine Learning capabilities, highlighting AutoML’s wizard-style model building and Designer’s drag-and-drop visual pipelines in the Azure Machine Learning Studio.

Explore the overview of computer vision and the four solution types—image classification, object detection, optical character recognition, and facial solutions—with core outputs like bounding boxes, confidence, and text.

Prioritize object detection and optical character recognition workloads for computer vision solutions, as the exam emphasizes these areas, including identifying the object's class, a probability score, and bounding box coordinates.

Provision and explore Azure AI Vision with Vision Studio in Visual Studio, create a computer vision instance, then analyze images with captions, object detection, and optical character recognition.

Learn how to optimize computer vision tasks with the general Azure AI service to cut costs, and use Custom Vision for training on custom data by uploading images.

Provision an Azure AI language instance and explore language studio to extract keyphrases, detect named entities, analyze sentiment, and present extractive and abstractive summaries.

Compare named entity recognition and entity linking in Azure AI language service for AI-900. Grasp utterance, entity, and intent; build a knowledge base for pre-built question answering from FAQs.

Provision an Azure document intelligence instance and explore document intelligence studio, using read, layout, and pre-built models like receipts to extract text, layout details, and structured data with confidence scores.

Discover Azure AI Search, a lucene-based service that indexes data into JSON. Indexing and querying workloads enable rich search, AI enrichment, and a knowledge store via skill sets.

Provision an azure ai search instance, configure data sources from blob storage, sql database, and cosmos db, create indexes and skill sets, and explore service features in the demo.

Identify features of generative AI models, including content creation, customization, interactivity, and versatility, explore prompt engineering, Copilot applications, and responsible AI principles (fairness, reliability, privacy, inclusiveness, transparency, accountability).